US3437713A - Process for separating olefins from admixtures with high porosity cuprous halide salts - Google Patents

Description

Apzli 8, 1969 5, LONG 3,437,713

PROCESS FOR SEPARATING OLEFINS FROM ADMIXTURES WITH HIGH POROSITY CUPROUS HALIDE SALTS Filed Sept. 21. 1967 Sheet of 5 ROBERT B. LONG lnvemor Patent Attorney April 8, 1969 R. B. LONG PROCESS FOR SEPARATING OLEFINS FROM ADMIXTURES Z of 5 WITH HIGH POROSITY CUPROUS HALIDE SALTS Filed Sept. 21, 1967 Sheet FIG. 2

ROBERT B. LONG lnvenlO Pcnent Attorney R. B. LONG April 8, 1969 PROCESS FOR SEPARATING OLEFINS FROM ADMIXTURES WITH HIGH POROSITY CUPROUS HALIDE SALTS Filed Sept. 21. 1967 Sheet FIG. 3

ROBERT B. LONG Inventor By ?M 4 MM Patent Attorney Apnl 8, 1969 R. B. LONG 3,437,713

PROCESS FOR SEPARATING OLEFINS FROM ADMIXTURES WITH HIGH POROSITY CUPROUS HALIDE SALTS Filed Sept. 21, 1967 Sheet 4 of 5 ROBERT B. LONG Inventor By #M 4M Patent Attorney Aprll 8, 1969 R. B. LONG 3,437,713

PROCESS FOR SEPARATING OLEFINS FROM ADMIXTURES WITH HIGH POROSITY CUFROUS HALIDE SALTS Filed Sept. 21. 1967 Sheet LARGE DEMONSTRATION UNIT (P05) 8 A G ,T C U D o R P RECYCLE PRODUCT GAS DECOMPLEXER 1 COMPLEX ER STRIPPER S A G N. M

S A G T C u D o R P E L C C E R RECYCLE RECYCLE PRODUCT TAIL GAS GAS Fig. 5A

ROBERT 8. LONG Inventor Potent Attorne' United States Patent Oflice 3,437,713 Patented Apr. 8, 1969 U.S. Cl. 260681.5 33 Claims ABSTRACT OF THE DISCLOSURE Improved novel processes for the separation of a compound capable of forming a complex with a cuprous halide from a mixture containing it are (a) separations other than diolefin from monoolefin separations effected in the absence of the presence of water or alcohols utilizing solid unsupported porous unitary particles of cuprous chloride or cuprous bromide having a porosity of above (of the total volume of a particle) 550l0,000 A. pores and (b) vapor phase separations generally including diolefin from monoolefin separations using said unsupported high porosity cuprous chloride or bromide particles. Also, preferred methods of preparing said high porosity particles are described.

CROSS REFERENCE This application is a continuation-in-part of U.S. patent application Ser. No. 601,208 filed Dec. 22, 1966, which is a continuation of US. patent application Ser. No. 333; 926 filed Dec. 27, 1963 'which is a continuation-in-part of U.S. patent application Ser. No. 163,075, filed Dec. 29, 1961, all now abandoned.

FIELD OF INVENTION This invention relates to processes for separating compounds wherein a mixture of compounds is contacted with solid particles of cuprous halide, at least one of the compounds reacts to form a cuprous halide complex, and the complex is then separated from the uncomplexed compounds and dissociated by heating to recover the complexed compound in high purity.

PRIOR ART The field of effecting separations and recovery of unsaturates, e.g. butadiene, ethylene, etc. with solid cuprous chloride or bromide has been exhaustively explored particularly over the past 30 years. During this period in addition to numerous publications some 60 patents have issued to seven major companies describing fixed bed, moving bed, fluidized bed, and slurry processes used with both supported (to increase surface area) and unsupported cuprous halide particles and with both liquid and gaseous feeds. For such processes to be commercially successful it is necessary that the cuprous halide reagent be (a) sufficiently active for substantially all the complexing material to react in a short feed residence time (to mini mize valuable material left in the treated feed and size of reactors respectively) and (b) sufficiently accessible for a high percentage of the theoretical capacity (27 wt. percent on CuCl for butadiene) to be reacted (to minimize adsorbent circulation between adsorption and desorption and attendant cooling and heating loads).

With respect to achieving both high activity and capacity with unsupported cuprous halide complexing agents, the patents e.g. U.S. 2,973,396 describe only the importance of the fineness of the state of subdivision of the particles and the purity of the cuprous halide both of which, although providing improvement, are orders of magnitude improved by the preparation of the present invention.

A series of patents does disclose methods of improving the activity of complexing agents of the cuprous halide impregnated on a non-adsorbent carrier type but they do not suggest that any of these are useful to improve the activity of unsupported cuprous halide complexing agents. Thus, U.S. 2,386,354 disclose multiple impregnation to obtain high surface area exposure due to ultra thin layers of cuprous halide in the pores and on the surface of the supports, U.S. 2,401,114 discloses cuprous chloride pow der uniformly distributed on shredded asbestos fiber in a relatively loose but cohesive mass and U.S. 2,756,267 states that the above preparations sometimes produce high activity reagents and sometimes do not for unknown reasons and that improvement is obtained by dissolving cuprous halide in an inert (preferably a monoolefin) solvent, adding a diolefin to precipitate the diolefin cuprous halide addition compound which is separated and dryed and then mixing with sawdust preferably containing an outside film of oil, and heating to decompose the complex and thus prepare the improved complexing agent.

Another group of patents discloses certain liquid phase diolefin separation processes which inherently produce at some stage in the process high porosity adsorbent. However, this was unappreciated by the patentees and there was no suggestion of removing the particles and using them externally in a different process. Typical of these diolefin recovery patents are (a) aqueous precipitation and slurry U.S. 2,386,333, U.S. 1,988,479, (1b) hydrocarbon precipitation and slurry U.S. 2,386,356, U.S. 2,395,- 955, U.S. 2,359,020.

SUMMARY The present invention is the discovery that solid unsupported particles of cuprous chloride or cuprous bromide can 'be prepared having a highly porous interconnecting large pore structure (e.g. 35% of the total volume of the particle 550-10,000 A. pores) allowing ready access of the feed to particle surface. Also, that these particles used in novel separation processes provide the high activity and capacity needed for commercial processes. It is noted that in addition to the discovery that prior art contacting of solid cuprous halide with liquid phase crude diolefin produces activation that it has now been discovered that the following do not produce the present activation: (a) liquid phase contacting with diolefins and other ligands complexing in above 1:1 stoichiometry in the absence of monoolefin, alcohol or water activators, b) liquid phase contacting with 1:1 ligands with or Without activators and (c) vapor phase contacting generally i.e. with any ligand with or without activators. The present invention also embraces discovery of various preparation procedures particularly adapted to produce superior particles for fluidized bed use.

THE DRAWINGS The drawings, FIGURES l-4, present photomicrograph views of the active high porosity cuprous halide particles as compared to conventional purchased cuprous halide particles and FIGURES 5 and 5A present a flow plan for carrying out a vapor phase separation process in which the cuprous halide particles are suspended as fluid beds by the upflowing vapors.

The structure of the present new cuprous chloride and bromide particles is seen from the accompanying drawings which are reproductions of photomicrographs of typical actual samples. Referring to FIGURES 1 and 2 electron microscope pictures (magnification 10,000x) are presented of replicas (prepared by dusting particles on a polystyrene wafer, lightly pressing the wafer between two slides for minutes at 330 F., cooling gradually, dissolving the copper salt out with HCl, coating the imof the present particles. It is also noted that mercury porosimeter tests on 2040 11. glass beads indicate 0.002 cc./gm. of the desired 550/l0,000 A. pore volume (i.e. essentially none). This indicates that the 550/ 10,000 A. pore volume results for the new CuCl and CuBr particles pression side with germanium metal, coating with car- 5 are real and not measurements of spaces between partibon vapor and finally dissolving away the polystyrene to cles. The unique porous structure of the present new comleave a several hundred angstrom thin, carbon replica plexing agents permits entry of the complexing gases shadowed with germanium) of preferred particles of which swell the membrane (i.e. the walls of the inter- CuCl (FIG. 1) and CuBr (FIG. 2) prepared by comconnecting honeycomb pores) to approximately twice plexing CuCl or CuBr dissolved in HCl or I-lBr respectheir original thickness as complexing proceeds but withtively with butadiene, precipitating (growing) the parout closing these pores due to their large diameter at least ticles in water and dissociating. As can be seen the paruntil complexing proceeds to near theoretical capacity. ticles have a well defined highly porous structure, the It is noted that surface area measurements on the comparticular particles shown being about 10-l5 microns in plexed particles indicate approximately zero surface area. average diameter (smaller than optimum size particles It should be noted that the present complexing absorption shown so that shape and edges can be seen). Additionally, is a chemical combination with the membranes and not as can be seen the individual pores shown are e.g. about a capillarity effect filling the large pores as present in 1600 A.-3200 A. in width. A discussion of the presence e.g. a sponge. Similarly on decornplexing (which does not of very small pores which are also preferably present decrease, i.e. shrink the size of the particles) the unique and are seen by porosimeter measurements only, follows. open (interconnecting) porous structure of the particles It is of interest to compare this material with commerpermits rapid release of the gases without destroying the cially available C.P. cuprous chloride. Referring to FIG- said porous structure. The said open porous structure URE 3, an electron microscope picture of a replica (preoifers an additional advantage in preparation of high pared as previously described) of particles of such com- 2 purity product in that a small amount of decomplexing mercial cuprous chloride is presented for comparison. As prior to the main desorption has been found to be excan be seen these much more dense particles have an tremely eifective to obtain sweeping out of the non-sealmost completely smooth surface. It has been found this lectively adsorbed contaminants, e.g. acetylenes, absorbed is true also of 100% purity cuprous chloride prepared by on the surfaces of the pores mainly by the gases evolved recrystallizing CuCl from concentrated I-llCl, etc. from the complex. This produces two advantages: (1)

From the above described electron microscope pictures minimization of the amount of stripping gas required and of the surface of the present new cuprous chloride and (2) more important, since product gas from the desorber bromide particles and from the mercury porosimeter and has been found to be the preferred gas to be used for CCl porosimeter data on pore volume now presented stripping, minimization of contamination by the same showing that approximately 20-50% of the total volume residual nonselectively adsorbed impurities present is large size pores the extraordinary activity and capacity therein. It should be noted that this open access to the of these particles can be understood (it is noted that membranes of the highly porous large pore, large size, porosimeter measurements are presented in terms of the pure (unsupported) particles having the necessary equivalent diameter of round pores i.e. tunnels and will strength to permit the required circulation around cooling be higher than the widths howrron the photographs); surfaces (e.g. in a fluidized bed), is indeed a major step Pore Volume, ctr/gm.

CClr Hg Porosirneter Porosimeter,

s00 A. 70-5b0 A. o-10,0001t. l0,000 A.

CuCl of Fig. 1-- Nil 0. 007 0.197 0. 044 CuBr of Fig. 2 2 0. 04 0. 009 0.117 0. 083 01101 recrystall (similar to Fig. 3)-after use in Fluid Bed i.e. to develop maximum pores 0. 01 0. 006 0. 009 0. 187

1 This volume is in the 550-800 A. range.

The above data show that the present new cuprous chloride and bromide particles are unique in having enormous open porosity in the 55010,000 A. pore diameter range. It is noted that open (i.e. interconnecting pores) porosity only, is measured by porosimeter). Thus, converting the (cc./gm.) pore volume figures to (cc./ cc.) figures by dividing by density of the solid to get solid volume and adding pore volume and solid volume to get total volume of 1 gram of particles and finally calculating the fraction of such total volume as pores gives respectively for CuCl and CuBr 41.1% and 29.3% of the total volume as 550-10,000 A. pores. It should also be noted that relatively small amounts of pores below 550 A. are present (which as will be described is important to the activity of the particle both due to partial condensation effects and due to providing additional access between pores), and that the figures for pores of above 10,000 A. (Lu) are presented merely for completeness and are probably mainly measurements of spaces between particles and thus essentially meaningless with respect to the pore volume forward over prior art suggestions to allow such access by coating supports, e.g. distributing extremely fine powder on the surface, impregnating to obtain a monomolecular layer, etc. Use of supports of course reduces selectivity due to nonselective absorption on the support itself. It is noted that the surprising strength and retention of activity and capacity of the present new particles in vapor phase fluidized solids (fiuid dense bed) use is witnessed by laboratory data showing that at least 15 cycles (complexing-decomplexing) use of active CuCl on commercial piperylenes concentrate 10/ 30a particles precipitated from an acid solution of the complex) did not appreciably reduce activity and capacity or particle size. Additionally, laboratory fixed bed tests showed maintenance of activity and capacity for at least 2,000 cycles on technical grade ethylene and at least 225 cycles on C.P. butadiene. Although the above description of the mechanism producing the present superior process is believed to be correct, it is of course, not intended to limit this invention thereby, the particular new cuprous chloride and bromide particles and critical processes for their use being the invention taught.

The present new cuprous chloride and bromide complexing agents have the following physical characteristics:

(1) Size-above about 50%, preferably above about 65%, more preferably above about 80% by weight of particles 1-1600/L, preferably -600/1, more preferably -300/L, yet more preferably 300-200,u., most preferably 50150/L (average diameters). In all embodiments the particles are regular unitary (rigid continuously joined structures, not small particles physically aggregated by surface effects only) particles slowly grown from solution as complexed single particles or complexed macroparticles composed of continuously joined microparticles, i.e. dendritically grown, which retain after dissociation within the porous structure the single particle or macroparticle composed of continuously joined microparticles structure (rather than agglomerates of small particles rapidly nucleated from solution and then physically combined by surface attractions only i.e. typical agglomerates have considerably less strength than the present invention particles). Examples of especially preferred particles are: (1) particles above 50%, preferable above 80% 10-50 single particles or macroparticles composed of up to four continuously joined above 5-30 length microparticles and (2) particles above 50%, preferably above 80% 50-400 preferably 75-300,!L, more preferably 75-150 u spherical macroparticles composed of more than four continuously joined microparticles, the said microparticle being of about 530/L (length) in size. It is noted that X-ray measurements show a very small 400-1300 A., usually 600-1100 A. basic crystals size within the single particle or macroparticle composed of continuously joined microparticles. This basic crystal size was found to be in general identical regardless of the preparation technique. It is believed that the Wall of the membranes separating the pores are thus probably only several basic crystals thick, i.e. from the e.g. 2000 A. Wall thicknesses shown on the electron micrographs. It is remarkable that open porosity sponge-like particles are so crystalline and have such unusual strength in e.g. fluid bed use. The large spherical macroparticles are particularly desirable since the smooth spherical structure has been found to provide excellent attrition resistance e.g. in fluid bed use and in addition it has been found that the present large 100-300 t spheres provide excellent smooth fluidization over a wide range of gas rates along with high activity and capacity. It should be noted that despite the fact that these smooth spheres are macroparticles of continuously joined microparticles (filleted to provide a smooth sphere surface rather than the sharp corners of the single crystals), the open porosity for complexing throughout the crystal persists (shown by porosimeter). A microscopic picture (magnification 60X) of the preferred 100-300 spheres prepared in Example 2 is shown in FIGURE 4. The individual microparticles making up the spherical macroparticles can be seen in the picture. The filleting growth between microparticles obtained in the growth of the smooth spheres can also be seen. A further proof of the (continuously joined) unitary rigid structure of the present porous particles both small single particles and macroparticles, and large spherical macroparticles is shown by the fact that the complexed particles have enormous strength, e.g. it has been found that complexed particles cannot be broken between two microscope slides by strong hand pressure. Typical normal aggregates even though complexed would be broken by such hand pressure.

(2) Porosity-above 10%, more preferably above 15%, yet more preferably above most preferably above of total volume of the particle pores of 550/ 10,000 A. diameter, preponderantly above 2000 A. diameter. Also, preferably particles have 01-15%, preferably 0.3-5 most preferably 0.5-3% of total volume of pores of 1-550 A., preferably 70-550 A. (small pores permit condensation but amount is small enough so as not to limit activity on complexing or strippability (for product purity) due to diffusion limitations). It should be noted that the present large pore porosity is unusual in that conventional porous solids such as cracking catalyst, catalyst supports, adsorbents including molecular sieves, all have the bulk of their pore volume smaller than 100 A.

(3) Puritypreferably above CuCl or CuBr, more preferably above yet more preferably above 97.5%, most preferably above 98.5% (the higher purities are desirably not only to obtain high activity and capacity in use, i.e., maintenance of high large pore porosity but also to increase mechanical strength and to prevent polymerization and polymer laydown on the particles in sustained multicycle use).

(4) Surface areapreferably above 2 m. /gm., e.g. 3-20 mF/gm. It is noted that the higher surface areas within the above range are not per se preferred because of stripping problems associated with higher surface areas. In preferred embodiments the gross morphology of the particles has been found to be either short, rectangular or hexagonal cross-section needles (preferred for the single macroparticle embodiment), spheres (preferred for the dendritic macroparticle ebodiment), pyramids, or cubes depending on the method of preparation. These particles have approximately equal height, width and length.

The new cuprous chloride or bromide particles of this invention are prepared by the steps of slow precipitation (growth) of crystalline cuprous chloride or bromide diolefin, acetylene, nitrile, or carbon monoxide complex particles from a liquid containing the complex in solution whereby solid cuprous chloride or bromide complex particles above 1a, preferably above 10a average are formed and dissociating said precipitated solid complex to obtain after dissociation the highly active porous cuprous salt. In general any method may be used so long as uniform solid complexed particles slowly precipitate (i.e. are form- 1 ed or grow) from solution in a size above about 1 preferably above 10;. Thus, it has now been found (as will be demonstrated in Example 21) that slow rates of macroparticle growth are required to produce the desired large (above 10a, and other preferred size ranges recited above) strong porous particles but that the permissible growth rate depends both the ligand used to make the complex and on the solvent from which the complex is grown. Average rates of crystal growth over the period of precipitation, i.e. grams/hour/liter are determined as follows: grams of complex recovered is divided by the time (hours) over which precipitation occurred and the resulting figure is divided by the total liters of original cuprous salt solution (a fuller discusion of this rate is contained in Example 21). The following slow rates are highly preferred for growth of good above 10a complex particles from solution, especially inorganic acidwater solutions: 1000 gms./hour/liter, preferably 500 gms./hour/liter, more preferably 5-300 gms./hour/liter, most preferably (for butadiene complex particularly) 25-100 gms./hour/liter. It is noted that rates less than 5 gms./liter/hour may be used but that the particles tend to be too large and of course such long reaction times unnecessarily increase the cost of preparation. While these rates are averages over a steady rate addition period, they are critical at the beginning of the period to prevent excessive nucleation. Higher rates may be used after the first 10% of precipitation has occurred. For more soluble (than butadiene) complexes (Example 21) of other complexing materials precipitation rates can be increased 3-fold. The preparation of preferred large above 50y, preferably 50-150 spherical macroparticles composed of continuously joined microparticles is obtained by using particularly slow precipitation rates especially at the beginning of precipitation to prevent excessive nucleation. For butadiene as the complexing material it is particularly preferred to add the antisolvent e.g. water to the solvent e.g. concentrated HCl to obtain a slow decrease in solubility and to use precipitation rates of less than 100 gms./hour/liter. With complexing materials which produce more soluble complexes higher rates may be used, still preferably below 1000 gms./hour/liter, but it is still preferred to decrease solubility slowly.

The above slow rates are obtained by the following general expedients which by creating a condition of low supersaturation for a considerable length of time during the precipitation either favors growth of initially formed particles or suppresses nucleation of new particles or both: (1) controlled slow rate of addition (mixing) of the particular reagent(s) which cause the cuprous chloride or bromide complex to precipitate out, or (2) controlled slow change in conditions to create low supersaturation. Specific expedients falling within (1) and (2) above which may be used alone or in combination are: (a) slow decrease in solubility by slow addition of antisolvent to the solution of the complex (rather than addition of solution to the antisolvent), (b) slow addition of the complexing material to solution of CuCl or CuBr (preferably low concentration) in a solvent in which the complex is less soluble than the CuCl or CuBr, (c) slow change in temperature to decrease solubility, (d) slow evaporation of solvent, (e) a controlled slow change in conditions during the latter part of the total precipitation period to create low supersaturation, e.g. slow addition of the solution of the complex to a limited amount of the antisolvent, although initial precipitation is rapid, as the amount of solvent builds up in the antisolvent low supersaturation occurs and the particles grow (well known percolation growth). All of these methods suppress nucleation of new particles and favor growth on nuclei already present.

It should be noted that care should be taken to prevent decomposition of the complex during precipitation by utilizing the well known temperatures and pressure within which the complex is stable, e.g. either low temperatures or high partial pressures of the complexing material or both may be maintained during precipitation.

Examples of general methods of preparation which by using the techniques recited above produce the large complex particles from which the desired new porous chloride or bromide is obtained upon dissociation are:

(1) Slow precipitation of solid complex particles from a liquid containing the complex in solution by either (a) addition of antisolvent or (b) change in temperature to decrease solubility, or (c) evaporation of solvent;

(2) Slow precipitation of solid complexed particles from uncomplexed cuprous chloride or bromide in solution in a solvent having a lower solubility for the complex than for the uncomplexed CuCl or CuBr by either (a) addition of a diolefin, acetylene, nitrile, or carbon monoxide to eg dilute aqueous HCl or HBr or (b) addition of a diolefin, acetylene, nitrile, or carbon monoxide, and an antisolvent e.g. water;

(3) Slow precipitation of complex from uncomplexed solution of a different copper salt (e.g. cupric chloride or bromide dissolved in water) by either (a) addition of a diolefin, acetylene, nitrile, or carbon monoxide and a reactant to convert the different salt to cuprous chloride or bromide, e.g. a reducing agent or (b) addition of the reactants recited in 3(a) plus an antisolvent; and

(4) As in (1), (2) and (3) except that a solid copper salt, preferably cuprous chloride or bromide having the large particle sizes recited above for the final new particles, is suspended (slurried) in a liquid mixture of a diolefin, acetylene, nitrile, or carbon monoxide and an activating material selected from the group consisting of monoolefin solvents for the CuCl or CuBr, alcohols, glycols and water and mixtures thereof and in which the desired diolegn, acetylene, nitrile, or carbon monoxide complex is less soluble than the copper salt, the said mixture having a slight solubility for the CuCl or CuBr so as to create a driving force whereby the diolefin, acetylene, nitrile, or carbon monoxide and activating material cause a phase boundary between cuprous salt and cuprous salt complex to move through the particle to in effect grow a new complexed particle. This new particle upon dissociation has the desired large pore pore structure.

It should be noted that in some embodiments of the preparations (1), (2) and (3) above, the complex is in solution only instantaneously and that the solid complex particles begin to nucleate and grow from the liquid complex immediately. This, however, does not interfere with the growth of the desired porous structure particles of relatively large size of this invention since under the proper conditions recited above growth is still favored over nucleation.

In each of the above methods any solvent for the copper salt may be utilized, i.e. organic or inorganic, the choice being dictated by the relative solubility of the cuprous chloride or bromide complex (more insoluble) vs. the solubility of the uncomplexed cuprous chloride or bromide. Suitable solvents for cuprous chloride or cuprous bromide are e.g. aqueous concentrated inorganic acids or aqueous dilute inorganic acids, (preferably the corresponding 2 to 12 normal HCl or HBr), and very concentrated aqueous inorganic salt solutions. Suitable solvents for cupric salts are e.g. water, alcohols, glycols, aqueous salt solutions, e.g. NH Cl, NaCl, KCl. It should be noted that a solvent should be chosen which does not contaminate the butadiene, acetylene, nitrile, or carbon monoxide complex of the cuprous chloride or bromide, i.e. the precipitate must not be contaminated, e.g. addition of butadiene and a reducing agent to cupric chloride in H O, etc. causes precipitation of uncontaminated complex due to the complete solubility of the cupric salts in the solvent.

Suitable antisolvents are any nonsolvents for the complex which are at least partially miscible with the primary solvent, e.g. water, C to (I and higher alcohols, e.g. methanol, ethanol, isopropanol; water soluble ethers, e.g. diethyl ether, water soluble ketones, e.g. acetone and methyl ethyl ketone; water soluble esters, e.g. methyl acetate; caustic solutions, e.g. sodium hydroxide solution. It is noted that the term antisolvent" includes materials which react with a solvent, e.g. sodium hydroxide to change the solvent medium.

It is noted that some solutions will involve partial or complete formation of a complex with the solvent. This will not present a problem so long as the solvent complex is less stable or more soluble than the desired solid 'butadiene, acetylene, nitrile, or carbon monoxide complex.

Preferably, unless the precipitation system involves extremely selective precipitation it is desirable to start with high purity cuprous chloride or bromide to prevent entrapment in the crystals of crystal lattice impurities which reduce the activity of the particles in the commercial complexing-decomplexing use of the particles, i.e. such impurities are barriers to movement of the ligand from the pore surface into the crystal lattice. Thus, it has been found that small amounts of impurities, e.g., 5% or above in the cuprous chloride or bromide produced greatly decreases activity, i.e., much more than could be attributed to the effect of this percent of inert material. Aside from effect on activity it has been found that cupric impurities cause contamination of the product by formation of organic chlorides.

It is also desirable to avoid contamination of the surface of the particles with diolefin, acetylene, or nitrile polymers or reaction products of these materials (or of diolefin, acetylene, nitrile, or carbon monoxide) with the solvent, produced in the preparation process, e.g., acid solvents in particular tend to react with unsaturates, etc. or cause condensation to form polymers. Preferred methods for limiting such contamination include utilizing low temperatures, limiting the amount of diolefin, acetylene, nitrile, or carbon monoxide supplied to only that amount needed to form the complex, removing polymer, etc. formed by filtration, decanting, washing, etc.

Suitable complexing agents which can be used as described to prepare the active cuprous chloride and bromide of this invention are any normally gaseous or liquid complexing compounds which form a stable complex having a ratio of copper to complexing compound greater than 1, preferably 2 or more. Such compounds are those having more than 1 pi bond per molecule. Such compounds include both materials which form only complexes having said ratios of copper to complexing compound greater than 1 and compounds which form complexes having a ratio of 1 or less which upon decomplexing pass through a stable complex having a ratio of copper to complexing compound greater than 1. Thus, the present inventor has found experimentally that certain materials, e.g. nitriles, diolefins, acetylenes, carbon monoxide under ordinary conditions forming a 2:1 complex can be made to complex in ratios of copper to complexing compound of 1:1 or less. However, upon dissociation complexing material is released selectively from the bed of cuprous chloride or bromide until the stable above 1:1, i.e. 2.:1 stoichiometric complex is completely formed before further decomplexing to the uncomplexed cuprous chloride or bromide occurs. In this specification by stable complex is meant a stoichiometric complex stable upon dissociation as described in the preceding sentence. As will be further discussed in the examples in connection with the experimental data it is the stable complex having a ratio of copper to complexing material of above 1:1 from which the large pores develop upon dissociation (due to the bonding of one molecule of the complexing material to more than one copper atom). It is noted that wherever in this specification diolefins, acetylenes, nitriles, or carbon monoxide are mentioned in connection with the preparation technique it is intended that these other compounds can be used. Preferred materials are carbon monoxide, HCN, and C -C or higher organic compounds containing at least one of the following functional groups:

/ (polyolefins) (2) CEC or (3) CEN and mixtures of these, wherein R is C or an alkylene group. More than one of these functional groups may be present in a single molecule. In addition other functional groups may be present 50 long as these do not interfere with complex formation. Preferred materials are C to C or higher, preferably C to C conjugated or nonconjugated aliphatic, cyclic or alicyclic diolefins, or less preferably polyolefins, e.g. allene, butadiene, isoprene, piperylene, octadienes, cyclohexadiene, cyclooctadiene, divinyl benzene, cyclododecatriene; C to C or higher, preferably C to C aliphatic or alicyclic acetylenes or acetylenes containing additional unsaturation, acetylene, methyl acetylene, propyl acetylene, phenyl acetylene, vinyl acetylene, etc.; and C to C or higher, preferably C to C aliphatic or alicyclic saturated or unsaturated nitriles, e.g. acetonitrile, acrylonitrile, propionitrile, phenylnitrile, methacrylonitrile, ethacrylonitrile, etc. Pure streams or dilute streams (diluted with an inert gas or natural dilute petroleum streams, e.g. butadiene diluted with butene and butanes) can be used so long as the diluene does not interfere with the precipitation of the desired solid complex.

Reaction conditions for the above described precipitation or growth of the desired new particles of this invention are in general the well known temperatures and pressures within which the desired complex is stable. In general suitable stable temperatures and pressures varying with the particular complex are in the range of -80 to 80 C., preferably --20 to 30 C., and pressures are in the range of 1 to 165 p.s.i.a., preferably to 50 p.s.i.a. In general higher pressures and lower temperatures within the above ranges as is well known stabilize the complexes.

Examples of methods of preparation and preferred conditions to favor growth of relatively large, strong, uniform, unitary particles are as follows:

(1) Precipitation from solvent containing the cuprous complex in solution by addition of an antisolvent The liquid cuprous complex may be produced in a variety of ways. One technique for preparing the liquid cuprous chloride or bromide complex is to bubble gaseous diolefin, acetylene, nitrile, or carbon monoxide through an aqueous solution of the cuprous salt. Other known means for forming the particular liquid complexes are also useful. As an example of a preferred procedure, butadiene gas may be bubbled through a saturated solution of cuprous chloride in concentrated HCl until essentially no more butadiene is absorbed thus forming an aqueous acidic solution of the CuCl-butadiene complex.

In lieu of employing gaseous diolefins, acetylenes, or nitriles, to complex with the cuprous salt in solution, liquids may be used. As an example liquid butadiene, in excess, may be contacted with an HCl solution saturated with CuCl, and with good agitation similar results are produced. It is to be understood that the manner in which the liquid complex is formed is a matter of choice depending largely on the equipment and materials available. Preferably to prevent polymer formation and formation of chlorinated diolefin, acetylene, or nitrile materials the following conditions are utilized: Acid concentration low, cuprous salt concentration high, temperature low, use of stoichiometric amount only of the diolefin, acetylene, nitrile, or carbon monoxide, filtration or settling to remove polymer and other contaminants prior to adding antisolvent.

To effect precipitation the precipitating medium, etg. water, is preferably added slowly, e.g. dropwise, to the cuprous chloride or bromide complex containing solution. Large regular crystals are thereby obtained. A somewhat less preferred alternate comprises adding the CuCl or CuBr complex solution slowly, e.g. dropwise, to a reasonably large quantity of the precipitating medium. A refinement of this step involves spraying a mist of the liquid complex solution into the precipitating medium. By adding the liquid complex in a concentrated form to the precipitating medium an explosive crystallization is effected whereby the solid particles initially obtained are reasonably small in average particle size but grow as addition continues. In either preparation preferably the solutions are purged with the diolefin, acetylene, nitrile or carbon monoxide during precipitation to prevent any decomposition of complex.

In either preparation concentrations of CuCl or CuBr in solution are preferably 1 to 60 wt. percent, more preferably 5 to 25 wt. percent. Time for addition to effect precipitation is preferably /6 to 10 hours, more preferably 1 to 5 hours. The amount of precipitation medium (i.e. antisolvent) is preferably 0.2 to 10, more preferably 1 to 5 vol. per vol. of complex solution.

The precipitated crystalline product is then separated from the solution and preferably dried with isopropyl alcohol or by any conventional drying means before dissociation. The dried complex is preferably maintained in a non-oxidizing atmosphere in order to avoid formation of cupric salts. The precipitated cuprous salt complex is then dissociated under controlled conditions to provide the present new cuprous salt which is many times more active than the initial cuprous salt employed in preparing the liquid complex. The conditions required for dissociation will obviously depend on the specific complex employed since each has its own dissociation pressure curves shown in the literature and in co-pending application Serial No. 115,684, filed June 8, 1961, now US. 3,206,521.

(2) Precipitation from uncomplexed cupric chloride or bromide in solution Cupric chloride or bromide is dissolved in a solvent,

e.g. water or alcohol, and a liquid or preferably gaseous diolefin, acetylene, nitrile, or carbon monoxide is mixed with the solution while at the same time a reducing agent is added. The reducing agent, e.g. S or Na SO is preferably added very slowly so as to produce cuprous salt slowly enough so that the formation of the complex does not deplete the solution of supersaturation with the diolefin, acetylene, nitrile, or carbon monoxide. Preferably, the reducing agent is added in a dilute solution miscible with the solvent for the cupric chloride so as to favor growth of large particles. Where a gaseous reducing agent is used it is preferably diluted with an inert gas. It is also noted that the cuprous chloride formed is usually and preferably more soluble than the complex, thus also favoring growth of the complex particles. Time for addition of reducing agent is preferably A; to 10, more preferably 1 to 5 hours.

(3) Slurry of solid CuCl or CuBr with liquid complexing material and an activating agent Solid cuprous chloride or bromide is suspended (slurried) in a liquid mixture of a diolefin, acetylene, nitrile, or carbon monoxide and an activating material selected from the group consisting of monoolefin solvents for CuCl or CuBr, alcohols, glycols, and water and mixtures thereof and in which the CuCl or CuBr is more soluble than the complex, the said mixture having a slight solubility for the CuCl or CuBr so as to create a driving force whereby the diolefin, acetylene, nitrile, or carbon monoxide and activating material cause a phase boundary between cuprous salt and cuprous salt complex to move through the particle to in effect grow a new complexed particle. This new particle upon dissociation has the desired large pore porous structure.

Suitable activating agents for this slurry technique are C -C preferably C -C monoalcohols and 0 -0 preferably C -C glycols, water, dilute acids preferably halogen acids, and C to C branched or straight chain olefins having appreciable solubility as pure materials for the cuprous salt. Particularly preferred monoolefins are butene-l, isobutylene, pentene-l and hexene-l because of their high solubility for CuCl and CuBr. Particularly preferred alcohols are methyl, ethyl and n-propyl. Particularly preferred glycols are ethylene and propylene glycol. It should be noted that the activating material and complexing material must be present in the liquid phase in the slurry operation. Thus, it has been found that gas phase operations (e.g. crude butadiene which contains large amounts of isobutylene and butene-l) even conducted near the dew point so that trace condensation occurs on the commercial CP CuCl or CuBr is not effective to produce the new large pore porous CuCl or CuBr of this invention (see Example 1B). It is noted that operation at the dew point for long periods of time is equivalent to slurry operations. It is further noted that the activating agent may be dissolved in the liquid complexing agent or the complexing agent may be dissolved in the activating agent.

The amount of activating agent used is not critical. Small amounts or large may be used although to some extent more large pore porosity is obtained with larger amounts. Thus, any amount up to 90 wt. percent or more based on the total liquid mixture, preferably 10 to 50 wt. percent may be used. It is also especially preferred to start with relatively pure CuCl or CuBr preferably above 90 wt. percent, more preferably above 97 wt. percent, since contaminants in the crystal seem to block development of large pore porosity upon decomplexing besides also decreasing mechanical strength of the CuCl or CuBr produced.

It is preferred to start with CuCl or CuBr of the same particle size as the preferred final particle size described above, but preferably of a particle size greater than that ordinarily commercially supplied, although somewhat larger sizes may be used. It has been found that the initial and final particle sizes tend to remain the same probably because a single particle becomes complexed by a phase boundary between cuprous salt and cuprous salt complex moving through the particle. It is noted that smaller sized initial particles may also be used where large amounts of solvent are used and if sufficient time for growth of particles from CuCl or CuBr complex in solution (from redissolving of other particles) is permitted. However, this really falls within the other preparations, i.e. precipitation from solution described above.

The time required for slurry contacting to produce the new particles of this invention depends to some extent upon the amount of solvent employed, temperature, etc. This time can be easily determined experimentally. However, in general it is preferred to effect contacting for 0.1 to 20 hours, preferably 0.5 to 5 hours.

It is noted that in a much less preferred embodiment some large pore porous structure and some improvement in activity, etc., over prior art CuCl or CuBr can be obtained by liquid phase slurrying as above described in the absence of a solvent.

It is noted that although both the cuprous chloride and the cuprous bromide new complexing agents of this invention used in commercial complexing separations provide real improvements over prior art material, the new cuprous chloride complexing agent is preferred. Thus, the new cuprous chloride as compared to the new cuprous bromide, complexes at higher temperatures and/ or lower pressures and additionally has higher activity (shorter gas residence times may be used).

It is preferred to dissociate the complex both in preparation of the active particles and especially in multicycle use in effecting commercial separations in the substantial absence of liquids, i.e., to efficiently strip or wash the particles of liquids including those wetting the surface and pores before effecting dissociation of the complex in the decomplexer and also in the prestripper where partial dissociation is used to obtain higher product purities. This is necessary because it has been found that liquids having some appreciable solubility for the complex, present during dissociation, tend to anneal the large pores and thus reduce activity. Alcohols have been found to be particularly deleterious and liquid monoolefin solvents such as butene-l, isobutylene, pentene-l, and hexenel also should be excluded. It is noted that some of these such as isobutylene and butene-l tend to be completely stripped in raising temperatures for dissociation but it is still preferred to use care to obtain essentially complete removal before the dissociation or partial dissociation stripping step. Other deleterious materials are liquid nitriles and water. It is noted that the dissociation step appears to be critical, i.e., the main annealing occurs when these liquids are present during dissociation and little annealing occurs at other stages in the preparation or commercial use of the particles.

Finally and of very great importance it is preferred to conduct complexing in the vapor phase at a temperature within 15 C. preferably within 10 C., more preferably within 5 C. of the dew point. It has been found that with the new high large pores (550/ 10,000 A.) porosity particles which also contain small amounts of small, e.g., 70500 A. pores, these small pores permit partial condensation in these pores and nucleation occurs to obtain much higher capacities and activities than are otherwise obtained.

The present new cuprous chloride and bromide particles may be used in effecting more economic separations of any compound capable of forming a complex with cuprous chloride or cuprous bromide. Thus, this includes all the separations described in the voluminous prior art previously referred to and additional compounds which it has been discovered complex with cuprous chloride and bromide. Preferred materials which complex with cuprous chloride or bromide are inorganic materials such as carbon monoxide and organic materials containing up to about 16 carbon atoms, preferably up to about 12 carbon atoms, more preferably up to about 8 carbon atoms. The higher boiling materials can be complexed in the vapor phase by techniques such as the use of vacuum, carrier gases, etc. Any materials may be used as carrier gases which do not interfere with the complexing reaction, e.g., inert gases, organic or inorganic materials. Examples of preferred materials which complex with cuprous chloride or bromide are C -C preferably C -C more preferably C -C compounds having one or more of the following functional groupsthrough which the complex is capable of being formed:

Carbon monoxide is suitable for use as the ligand. Additionally, unsaturated carbonyl compounds such as propenal, butenal, pentenal, and the like; the various unsaturated ketones such as l-butene-R-one, 1,4-pentadiene-3- one, 2-pentene-4-one, and similar ketones may be employed. In general the alkane nitriles such as methane nitrile, ethane nitrile, propane nitrile, and higher nitriles are useful. Aryl, alkaryl and arylalkyl nitriles also complex with cuprous salt and may be used to form the liquid complex precursor. Unsaturated nitriles, such as acrylonitrile, methacrylonitrile, and ethacrylonitrile are further examples of ligands suitable for use in the present process. Ligands having a combination of functional groups selected from the list recited above are less preferred alternates. Also, other functional groups may be present so long as these do not interfere with complex formation.

Examples of olefins are ethylene, propylene, butylene, isobutylene, pentenes, etc. While alpha, non-alpha, straight and branched chain olefins are all employable, alpha olefins appear to complex more readily, presumably due to the absence of steric hindrance and are preferred. Di and triolefins such as propadiene, butadiene, isoprene, dicyclopentadiene, cyclopentadiene, octadiene, cyclododecatriene and the like, readily complex. Olefinic aromatic compounds such as styrene and the like may also be employed. The acetylenes such as methyl, ethyl, vinyl, propyl acetylenes and the like, as well as acetylene per se are also useful as ligands. It should be noted that compounds containing functional groups in addition to the functional group(s) through which the complex is formed may also be employed since they do not ordinarily interfere with complexing. Also, compounds containing more than one functional group through which the complex is capable of being formed may by proper choice of conditions (chosen based on the temperature pressure dissociation curve) be separated from another compound having one of the same functional groups, e.g., acrylonitrile from acetonitrile.

Complexing condilions may be chosen to be any temperature and pressure conditions under which a complex forms between the cuprous chloride or bromide and one of the compounds present in the mixture. These conditions may be chosen to effect vapor phase, liquid phase or mixed vapor liquid phase contacting with the cuprous halide. Similarly, decomplexing conditions may be chosen to be any temperature and pressure conditions under which the compound complexed dissociates, again vapor phase, liquid phase or mixed vapor liquid phase. In general for complexing temperatures are in the range of 80 C. to 100 C., pressures in the range of 0.5 to 125 atmospheres and residence times in the range of for feed 1 second to 2 hours, for cuprous halide minutes to 2 hours. Particularly preferred complexing conditions for liquid phase contacting of the cuprous halide are temperatures of 50 C. to 40 C., pressures of l to 100 atmospheres and contact times of 5 minutes to 2 hours. In general for decomplexing temperatures are in the range of -20 C. to 150 C., pressures in the range of 0.1 to 100 atmospheres and cuprous halide residence times of 1 minute to 2 hours.

Preferred new processes in one embodiment are processes for separating any of the compounds described above from mixtures with other compounds, except separations of diolefins from mixtures containing appreciable quantities (e.g. above 2 wt. percent) of monoolefins, carried out by contacting the solid highly porous cuprous chloride or bromide of this invention, under vapor phase, liquid phase or mixed vapor liquid phase conditions as described above in the absence of the presence of appreciable quantities (eg 1% or more) of water or alcohols.

In another embodiment preferred new processes are processes for separating any of the compounds described above from mixtures with other compounds including diolefins from monoolefins carried out by contacting the solid highly porous cuprous chloride or bromide of this invention under vapor phase conditions as described above.

A preferred apparatus for carrying out the preferred process (circulation of solids between two fluidized beds one operated on absorption and one on desorption) for utilizing the present new cuprous chloride or bromide particles is depicted in the accompanying drawing, FIGURE 4. Referring to the drawing vaporous feed is supplied through line 1 to complexer reaction vessel 2 containing a bed of cuprous chloride or bromide particles supported on distribution plate 3 and fluidized by the feed vapors. The reaction vessel is 2 ft. in diameter and contains approximately 100 1" OD. heat transfer tubes to remove the enormous heat released in Complexing. The distribution of these tubes is shown in the cross sectional view FIG. 4A (thru A-A) of the reactor vessel 2 showing the heat transfer tubes 4. As can be seen the free area is only approximately of the total area. Tail gas vapors leaving the fluid bed are passed overhead through line 5 to cyclone 6 from which entrained solids are returned to the reactor through line 7 and tail gas is removed through line 8. Desorbed particles are supplied near the top of the fluid bed through line 9 and flow downwardly through the fluid bed and bypass the distribution plate 3 through downcorners (not shown) to a stripper 4 where they are heated and/or stripped of nonselectively absorbed feed. Stripping gas may be supplied through line 10 or stripping may be effected primarily by heating. Particles are then passed upward through lift line 11 with lift gases being supplied through line 12 to the upper section of the fluid bed of decomplexer 13. This decomplexer also contains the approximately heat transfer tubes described in connection with the complexer vessel 2. Particles flow downward through the fluid bed in vessel 13 where they are fluidized by fiuidization stripping gas supplied through line 14, the particles again being supported by a gas distribution baflle 15. Overhead gases containing the desired product separated are passed through line 16 to cyclones 17 where solids entrained are separated and returned through line 18 to the reactor and the desired pure product gases are passed from the system through line 19 to storage. Desorbed cuprous chloride or cuprous bromide particles are passed through line 20 back to reactor 2 through lift line 21 with lift gases being supplied through line 22. In an especially preferred embodiment all of the stripping and lift gases are the product and tail gases respectively from the process. Thus, the stripping gas for the decomplexer and for the stripper and the lift gas to transport the CuCl or CuBr to the decomplexer is the high purity product gas (prevents contamination or dilution of the product gas). With respect to the lift gas used to supply decomplexed particles to the complexer this is the tail gas from the absorber. Thus, no additional impurities or dilution of the tail gas occurs. Of course, other stripping and lift gases, e.g. nitrogen, hydrocarbons, etc. can be used but these should be chosen so as not to interfere with complexing or decomplexing and so as to be separable, e.g. by distillation from the pure product recovered by complexing and also preferably from the tail gas.

In a preferred embodiment to obtain maximum product purity the particles withdrawn from the bottom of the absorber are slowly and uniformly heated as they pass downwardly through the stripper, the heat being supplied either by heat transfer tubes (not shown) or by heated stripping gas, preferably by both. Preferably residence times and temperatures are adjusted to obtain partial decomplexing of 110%, preferably 3-6%, of the compound complexed contained in the particles. The amount of stripping gas is preferably limited both to effect savings in stripping gas requirements and in the preferred embodiment where product gas from the desorber is used as the stripping gas to minimize recontamination of the particles by impurities contained in said stripping gas. Preferred amounts of stripping gas are 1-10 vol. percent, preferably 2-7 vol. percent, based on the amount of gas contained in the complexed particles passed to the stripper. In an additional preferred embodiment to obtain very high purity product the final stripping is conducted substantially with gases evolved from the partial decomplexing of the particles or with a minimum amount of additional stripping gas, preferably product gases, or more preferably a pure stripping gas such as nitrogen, to permit smooth flow of the particles. This latter embodiment provides maximum protection against contamination both by impurities in the stripping gas and by impurities in the gases evolved from the particles (impurities swept off the surface). The extremely high purities obtained in the separation of butadiene from refinery C streams (e.g. from steam cracking) by expedients such as described above are reported in Example 28.

Preferred complexing and decomplexing conditions used for effecting preferred commercial separations using preferably the above described system are as follows: (Preferred superficial velocity fluidization rates 0.05-5 .0, preferably 0.15-l.0 ft./sec.).

BUTADIENE SEPARATED FROM CRUDE BUTADIENE Preferred Most Preferred Complexing:

Temperature, C to 40. Pressure, Atmos 1 to 5. Gas Residence Time, secon 15 to 150 Solids Residence Time, minutes 20 to 100. Dccomplexing:

Temperature, C 100. 60 to 90 Pressure, Atmos 1 to 5. Gas Residence Time, seconds 15 to 150 Solids Residence Time, minute 10 to 100 ETHYLENE sEPARATEgRligROM STEAM CRACKING C2 ACRYLONITRILE SEPARATED FROM ACETONITRILE Most Preferred Preferred Complcxing:

Temperature, Cu-.." Pressure, atmos; Gas Residence Time, sc Solids Residence Time,

Dccomplcxing:

Temperature, C

0.5 to 10- 1 to 5. 1 to 400 15 to 150. minutes 10 to 200 20 to 100.

50 to 140 70 to 120. Pressure, atmos. 0.5 to 10". 1 to 5.

Gas Residence T seconds 1 to 400 15 to 150. Solids Residence Time, minutes to 200 to 100.

Norm-For the nitrile separations an inert carrier gas, such as N2, CH4, etc, must be used to prevent condens t on at the 1112 PIBSSHW$= CARBON MONOXIDE SEPARATED FROM HYDROGEN Preferred Most Preferred Complexing'.

Temperature, C -25 to 100... 10 to 60 Pressure, atmos r 0.5 to to 60.

Gas Residence Time, seconds 1 to 400 15 to 150.

Solids Residence Time, Minutes 10 to 200"." 20 to 100. Decomplcxing:

Temperature, C 20 to 140 40 to 120.

Pressure, atmos 0.5 to 100 1 to 60.

Gas Residence Time, seconds.-. Solids Resldence Time, minutes ALLENE SEPARATED FROM METHYL ACETYLENE Preferred Most.

Preferred Complexing:

Temperature, C 40 to 70.... 20 to 40.

PIPERYLENES SEPARATED FROM CYCLOPENTENE Preferred Most Preferred Complexing Temperature, C -10 to 80.- 0 to 70. Pressure, atmos 0.1 to 5 1 to 3. Gas Residence Time, seconds 1 to 400 15 to 150.

Solids Residence Time, minutes .I. 10 to 200".-- 20 to 100.

Docomplexing:

Temperature, C 40 to 60 to 110. Pressure, atmos 0.1 to 5 1 to 3. Gas Residence Time, seconds 1 to 400.....- 15 to 150. Solids Residence Time, minutes 5 to 200. 10 to 100.

N 0TE.TO operate with piperylenes at low temperature and the higher pressures a carrier gas such as nitrogen must be used.

The present invention will be more clearly understood from a consideration of the following examples and the laboratory data contained therein.

EXAMPLE l.HCl SOLN. OF CuCl-BUTADIENE COMPLEX ADDED TO WATER 8000 ml. of concentrated HCl and 100 g. of copper pellets were charged to a 12 liter beaker, sparged below the surface with CF. butadiene. 4472 g. of CuCl (which had been recovered from a slurry of equal Weights of concentrated HCl and Bakers C.P. CuCl which was stirred at 22 C. under a CO atmosphere in the presence of copper pellets to clean the surface) were then added and the mixture was stirred while continuing butadiene purge for about one hour to dissolve the cuprous chloride. Undissolved solids were filtered off and 4500 ml. of the saturated solution was transferred to a resin pot, 50 g. of copper pellets were added (to reduce any cupric salts present) and the mixture was purged with butadiene below the surface (with stirring) for about four hours at 30 C. until a second liquid phase appeared. The solution was then slowly added over a period of about 10 hours to five times its volume of distilled water also at about 30 C. saturated with butadiene and blanketed with butadiene. A light yellow precipitate of complex began to form as soon as the acid solution hit the water. The total precipitate formed was filtered off, and Washed with isopropyl alcohol and then diethylether to remove water and butadiene polymer from the crystals of complex. The ether was removed by blowing gaseous butadiene through the bed of complex crystals and the final product was bottled under a butadiene atmosphere. Analysis of the complex crystals showed that the crystals contained 100% of the theoretical amount of butadiene. The present invention highly active CuCl was obtained by dissociating these crystals. It is noted that porosimeter pore volumes and activity and capacity data in fluid bed tests along with other data on the active CuCl prepared in this example and in most of the following examples is presented in Example 20.

1 7 EXAMPLE 1A.HCl SOLN. OF POLYMER CON- TAMINATED CuCl-BUTADIENE COMPLEX ADD- ED TO WATER A saturated solution of cuprous chloride was prepared by adding 1935 g. of commercial C.P. CuCl and 100 g. of copper pellets to 6 liters of concentrated HCl at 22 C. and stirring for 30 minutes followed by filtering off the undissolved solids (540 g.). The saturated solution was then charged to a liter reaction flask and CF. butadiene (99%) vapor was added below the surface through a sparger tube. The flask was also provided with a Dry- Ice-alcohol condenser to reflux any unreacted butadiene to the CuCl solution. The butadiene was added over a period of 3 days and an insoluble liquid layer of over 1 liter formed on the top surface of the solution. The temperature during the entire addition of butadiene was maintained between 17 and 36 C. The oil layer was decanted and distillation and analysis of the cuts obtained from this upper oil layer showed that it contained large amounts of chlorobutenes, hydrochlorinated butadiene polymers as Well as other butadiene polymers. The entire remaining solution was filtered and 360 gms. of 1-4 mm. diameter tetrahedral uncomplexed CuCl crystals were obtained. The filtrate was divided into two parts and about six volumes of water was added slowly to one of them at room temperature. This resulted in the precipitation of 369 g. of unidentified greenish white semi-amorphous crystals which on heating to 175 C. yielded about 4 wt. percent of a liquid distillate, probably polymer, the remainder being relatively active CuCl (dissociated due to the heating step). This technique is very wasteful of both HCl and butadiene and presents a diflicult washing problem to remove the chlorinated polymers from product crystals. In addition the cuprous chloride is considerably less active than optimum. This example also shows that where the temperature is high and the time of butadiene addition is long in the complexing step excess formation of polymer and other contaminating by-products dilutes the acid causing uncomplexed cuprous chloride to precipitate out.

EXAMPLE 1B.-UNCOMPLEXED CuCl RE- CRYSTALLIZED FROM HCl A saturated solution of Bakers C.P. cuprous chloride is concentrated HCl (12 N) was prepared as described in Example 1 and the solution was added to six volumes of water at room temperature. The precipitate formed was filtered off and washed with isopropyl alcohol and then diethyl ether. The ether was recovered by blowing gaseous nitrogen through the crystals. The material was found to have very poor large pore porosity (from electron microscope replica picture), and capacity and activity in fluid bed tests. Thus, this example is presented to show that even the best cuprous chloride prepared absent the present invention, i.e. 100% pure cuprous chloride having the desired particle size for fluidization but without the des1red porosity is markedly inferior to the present invention material.

EXAMPLE 1C.EVAPORATED BUTENE-l SOLU- TION OF COMPLEX OF CuCl WITH BUTENE-l 9 /2 cu. ft. of gaseous butene-l were condensed at atmospheric pressure in a 2-liter resin flask cooled to 10 C. and 200 gms. of Baker and Adamson C.P. cuprous chloride was then added. The mixture was stirred for 2 hours and then undissolved material was allowed to settle out before decanting the clear cuprous chloride-butene-l solution to another cooled resin pot. B-utene-l was evaporated off at atmospheric pressure to crystallize out the CuCl-butene-l complex. When most of the butene-l had been removed and the solid complex was in the form of an essentially dry pale yellow powder it was transferred to a tared flask at Dry Ice temperature. The flask was allowed to warm up to room temperature and the gaseous butene-l that was evolved was metered through a wet test meter. After the first 3.5 liters of gas came off the cuprous chloride-butene-l complex was completely dry and broke into small pieces when the flask was shaken. Then an additional 5.9 liters of gas came off when the flask was heated to 35 C. The flask was reweighed after all the butene-l was removed and showed that 1 61 g. of cuprous chloride was present in the flask. Although the amount of gas evolved in the two steps measured was less than theoretical for the 1:1 olefin complex it is believed that this was clearly due to decomplexing occurring prior to such measurement and that the complex was initially substantially of theoretical. This material was found to have very poor large pore porosity and activity and capacity in fluid bed tests. Thus, this example is presented to show that materials complexing only 1:1 e.g. olefins are markedly inferior to the present invention material.

EXAMPLE 2.WATER ADDED TO HCl SOLUTION OF CuCl-BUTADIENE COMPLEX To 1300 ml. of concentrated HCl, 450g. of Bakers C.P. cuprous chloride were added along with 25 g. of copper pellets. Two ft. of butadiene were added to the solution in a stirred vessel over a 1-hour period and the temperature rose from 22 C. to 31 C. At the end of this period traces of oily polymer began to appear. One liter of water was then added dropwise over 2 hours and ten minutes, to the concentrated HCl solution purged with butadiene. An additional 9 liters of water were added over a subse quent 1 hours. An excellent cuprous chloride of average particle size of over 100 microns having a generally uniform size and spherical shape was obtained. This material was found to have excellent fluidization properties over an extremely Wide range of gas rates.

EXAMPLE 3.HC1 SOLUTION OF CuCl-ALLENE COMPLEX ADDED TO WATER The same procedure as that of Example 1 (except that no preliminary slurry cleaning of the surface of the cuprous chloride particles was conducted) was used on a. sample of commercial C.P. CuCl except that allene (1,3- propadiene) was used instead of butadiene. The yield of dried crystalline complex of allene with CuCl was some what less than for the butadiene complex and the final crystals contained 83% of the theoretical amount of allene.

EXAMPLE 4.ACETYLENE ADDED TO HCl SOLU- TION OF CuCl PRECIPITATES COMPLEX DI- RECTLY The same general procedure of Example 1 (except that the CuCl saturated solution was prepared by adding reagent grade C11 0 to the HCl) was used to effect precipitation except that acetylene was used instead of butadiene. Following addition of the acetylene a yellow precipitate was filtered off and the clear solution was added sloWly to 12 liters of water. A deep purple sludge was formed which turned black and a black powder was recovered which was found upon analysis to be CuCl and free carbon. The initial yellow precipitate found to be the desired complex of acetylene with CuCl was very much less than for the butadiene complex and the final crystals contained 50% of the theoretical amount of acetylene. This experiment indicates that diolefins are much preferred to acetylene but that acetylene may also be used.

EXAMPLE 4A.HCl SOLUTION OF CUPROUS CHLORIDE METHYL ACETYLENE COMPLEX ADDED TO WATER 1500 cc. of a room temperature saturated solution of Bakers analyzed cuprous chloride (copper pellets added to reduce any cupric chloride present) was prepared. This solution was purged with methyl acetylene and then 6000 cc. of water at 0.5" C. was added dropwise over a 2-hour period while purging with 5.2 ft. of methyl acetylene. No visible precipitate was formed until above 1500 cc. of Water had been added (cloudiness appeared). The final precipitate was recovered by filtration, washed twice with isopropyl alcohol and then twice with ethyl ether (in a methyl acetylene atmosphere) and the ether was finally removed by flufiing with methyl acetylene. 260 grams of the complex were recovered. Microscopic examination revealed the typical (for good activity) transparent complexed crystals which became opaque upon dissociation. They had a short rod shape.

EXAMPLE 5.-HBr SOLUTION OF CuBr-BUTADI- ENE COMPLEX ADDED TO WATER A similar procedure to that of Example 1 was used to prepare active cuprous bromide from a sample of B&A C.P. cuprous bromide (dissolved in HBr). The yield of dried crystalline complex of butadiene with CuBr was approximately equal to that of the butadiene complex and the final crystals contained 82% of the theoretical amount of butadiene.

EXAMPLE 6.LIQUID S ADDED TO ANI-IY- DROUS EtOI-I SOLUTION OF CuCl AND LIQUID BUTADIENE 40 g. of reagent grade Baker and Adamson cupric chloride were dissolved in 100 ml. of anhydrous ethyl alcohol, the solution was cooled to 25 C. and l5 ml. of liquid C.P. 1,3butadiene were added. 20 ml. of liquid anhydrous S0 was then added dropwise over a period of 5 minutes. A brownish yellow precipitate formed and the mother liquor was decanted olf. The precipitate was then slurried with 200 ml. of distilled water and became bright yellow. The precipitate was filtered oft, washed with 100 ml. of isopropanol, followed by 100 ml. of cold diethylether, and finally dried in a stream of 1,3-butadiene. Microscopic examination of the crystals showed the typical structure of the 2CuCltbutadiene complex. Analytical data showed 92% of the theoretical amount of butadiene in the crystals.

EXAMPLE 7.-Na SO WATER SOLUTION ADDED TO BUTADIENE PURGED \VATER SOLUTION OF CuCl 454 g. of Matheson, Coleman, & Bell cupric chloride (CuCl .2I-I O) were dissolved in 4000 ml. of water, the solution was cooled to C. and then the solution was purged with butadiene. 2000 ml. of an 18 wt. percent N21 SO water solution was added dropwise over a period of about 1 hour while continuing to purge with butadiene. A yellow precipitate was obtained which by microscopic examination was the 2CuCl2butadiene complex and which upon dissociation had the desired porous structure (electron microscope replica picture taken).

EXAMPLE 8.SO AND BUTADIENE GASES ADDED TO WATER SOLUTION OF CuCl 454 g. of Matheson, Coleman, and Bell cupric chloride (CuCl -2I-I O) were dissolved in 2000 ml. of water. The solution was cooled to C. and then purged with gaseous butadiene and S0 over a period of 4 hours. A yellow precipitate was obtained which by microscopic examination was the 2CuCl:butadiene complex and which upon dissociation had the desired porous structure (electron microscope replica picture taken).

EXAMPLE 9.-HCl SOLUTION OE CuCl-ACE- TONITRILE COMPLEX ADDED TO WATER A room temperature saturated solution of CuCl in concentrated HCl (12 N) was prepared by dissolving C.P. CuCl and filtering. 900 cc. of the solution was mixed with 900 cc. of acetonitrile. The mixture was added dropwise to 8 liters of H 0 containing 92 cc. of acetonitrile over a period of 2 hours. A white precipitate began to form as soon as the mixture was added. The product was filtered and washed with alcohol and ether and fluffed with nitrogen saturated with acetonitrile at room temperature. The product contained 29 wt. percent acetonitrile, i.e. 1:1 mol ratio and upon dissociation was found to have the desired large pore porous structure and high capacity and activity in fluid bed test. It should be noted that from other data obtained by the present inventor it is known that acetonitrile complexes in at least two stoichiometric forms of the complex which dissociate stoichiometrically, i.e. the less stable 1:1 complex dissociates completely to the 2:1 complex and only then to the decomplexed state. This was experimentally found to occur with the present particles, i.e. in a particle being dissociated under the microscope first a milky haze indicating very fine pores appeared and then the typical large pore material of this invention appeared i.e. opaque under transmitted light and solid white under reflected light due to the large pores appearing. By comparison non-porous particles are more transparent, and are glossy under reflected light. This example indicates that so long as the particles precipitated as the complex from solution on dissociation go through stoichiometric complex greater than 1:1, e.g. 2:1 the active large pore material is obtained. It should be noted that it will be shown in another Example, Example 19 that besides the necessity of utilizing a complexing agent which passes through a stoichiometric complex greater than 1:1 upon dissociation, it is also necessary to precipitate the complex from solution (or slurry in the presence of an activating material). It is theorized that two cuprous atoms are linked to a single molecule of acetonitrile in a stable state and probably only one of the same two copper atoms is linked to an additional molecule of acetonitrile in a looser state. The looser state bonding does not produce the large pore structure (i.e., similar to monoolefins bonding of Example lA) but it does not interfere with the formation of said structure (as shown in this example). Thus, as an over-simplified explanation of this, in all cases where the desired structure was obtained, more than one, e.g. two 1.- bonds of a single molecule of the complexing material are attached to more than one, eg. two copper atoms which multiple bonding upon decomplexing produce the large pores. By comparison where only a single 11' bond of the molecule is bound to one carbon atom the large pores are not obtained upon decomplexing. It is noted that by multiple bond we refer either to both double bonds of a diolefin or the two 1r bonds of acetylenes or nitriles or carbon monoxide.

EXAMPLE 10.HC1 SOLUTION OF CuCl-ACRYLO- NITRILE COMPLEX ADDED TO WATER The same general procedure as in Example 9 was used except that acrylonitrile was used instead of acetonitrile. The complex obtained contained 21.1 wt. percent acrylonitrile, i.e. 2 moles CuCl to 1 mol of acrylonitrile and upon dissociation was found to have a reasonably large pore, porous structure and high activity and capacity in fluid bed tests.

EXAMPLE 11.HCl SOLUTION OF CuCl-CARBON MONOXIDE COMPLEX ADDED TO WATER 1500 cc. of room temperature saturated solution of B&A C.P. QJCl was prepared. This solution was purged with nitrogen to remove any air and then 6000 cc. of water at 0-5 C. was added dropwise over a 2-2.5 hour period while purging with carbon monoxide. No visible precipitate was formed until about 2 liters of water had been added (cloudiness appeared). The final precipitate was recovered by filtration and had a silvery, flaky appearance. The precipitate was washed twice with isopropyl alcohol and then twice with ethyl ether (in a CO atmosphere) and the ether was finally removed by flufling with carbon monoxide. 260 grams of the complex were recovered. Microscopic examination revealed the typical (for good activity) transparent complexed particles which became opaque upon decomplexing. The particles were triangle shaped.

EXAMPLE 1lA.-Water added to HCl SOLUTION OF CuCl-AMMONIA COMPLEX cc. of room temperature saturated solution of B&A C.P. CuCl was prepared. The same procedure as in Ex- EXAMPLE 12.WATER ADDED TO H-Cl SOLUTION OF CuCl-METHYL ACETYLENE COMPLEX 1500 cc. of room temperature saturated solution of CuCl was prepared. The same procedure as in Example 11 was used except that gaseous methyl acetylene was used instead of carbon monoxide and that 6000 cc. of water was added dropwise over a period of 2 hours. A white precipitate began to form after addition of water. Fluffing was with methyl acetylene. The dissociated material had good large pore porous structure.

EXAMPLE l3.SLURRY OF C.P. BUTADIENE VOL. PERCENT METHANOL) WITH B&A SEG- REGATED CUPROUS CHLORIDE EXAMPLE 13A.SLURRY OF C.P. LIQUID PROPYLENE WITH CUPROUS CHLORIDE RE- CRYSTALLIZED FROM HCI A sample of the cuprous chloride recrystallized from concentrated HCl (added to water) was covered with liquid propylene (condensed in the reaction flask at 60= C.). The propylene-cuprous chloride slurry was stirred for about one hour and finally the propylene was boiled off by warming the complex to room temperature. A dry powdery product was obtained which had very low large pore porosity and poor activity and capacity. This indicates again that monoolefins cannot be used as the complexing agent to prepare the desired material.

EXAMPLE l3B.-SLURRY OF BAKERS C.P. CuCl WITH CONCENTRATED HCl CONTINUOUSLY PURGED WITH BUTADIENE 500 grams of Bakers C.P. CuCl (large excess) was stirred with 600 ml. of concentrated HCl in the presence of grams of metallic copper under a nitrogen blanket for two hours at room temperature. Butadiene was then substituted for N and purged below the surface of the liquid for four hours. A liquid polymer layer collected when the stirring was stopped. The slurry was allowed to stand overnight under a butadiene atmosphere and the solids were then recovered by filtration and washed with alcohol followed by ether and finally dried in a flowing butadiene stream. The dry powdery product had a very low large pore porosity indicating that concentrated HCl is not a particularly elfective activating agent for preparing the butadiene complex by the slurry technique.

EXAMPLE 14.4LURRY OF C.P. BUTADIENE (40 VOL. PERCENT METHANOL) WITH B&A SEG- REGATED CuCl 300 g. of Baker and Adamson segregated cuprous chloride were added to 400 cc. of methanol and the slurry was cooled to 10 C. Gaseous butadiene (7.4 ft. was then charged for 3 hours with stirring and the buta- 22 diene condensed in the reactor. The CuCl-butadiene complex after standing overnight was filtered off, washed with alcohol and then with ether and was dried in a stream of flowing butadiene. Upon dissociation the particles were found to have higher large pore porosity than where a smaller amount of alcohol was used. (of. Example 11).

EXAMPLE l5.-SLURRY OF C.P. BUTADIENE (73 VOL. PERCENT METHANOL) WITH BAKERS C.P. CUPROUS CHLORIDE 250 g. of Bakers C.P. cuprous chloride were added to 350 cc. of methanol and the slurry was cooled to 25 C. Gaseous butadiene was then charged for 30 minutes and the butadiene condensed in the reactor. A total of 146 g. of butadiene was added and the flask was stirred an additional 30 minutes after all the butadiene was in. The complex was filtered ofl and washed with alcohol then with ether and dried in a stream of flowing butadiene.

EXAMPLE 16.SLURRY OF C.P. BUTADIENE PLUS METHANOL WITH HCl RECRYSTALLIZED CUPROUS CHLORIDE This preparation was made in a fluid bed reactor at 25 C. by treating g. of concentrated HCl recrystallized (added to water) cuprous chloride with enough C.P. butadiene tocompletely cover the collapsed bed. Then 30 cc. of methanol were added. A nitrogen stream was then fed to the bottom of the slurry to stir it for about one hour. After one hour the bed was heated to drive off the butadiene and methanol and a flowing stream of nitrogen was used to fluidize and dry the bed. Upon dissociation the particles were found to have the desired high activity and capacity in fluid bed tests.

EXAMPLE l7.SLURRY OF CRUDE BUTADIENE WITH HCl RECRYSTALLIZED CuCl 200 g. of cuprous chloride recrystallized from concentrated HCl (added to water) were added to gms. of liquid crude butadiene from steam cracking (30.7 wt. percent butadiene, 52 wt. percent isobutylene plus butene- 1, the remainder being mainly paraflflns but also containing other butenes and acetylenes) at -60 C. and the slurry was stirred for one hour. The crude butadiene was allowed to evaporate oflf at the boiling point of the crude butadiene, i.e. about 6 C. The material was then decomplexed in the fluid bed with gaseous nitrogen and yielded an amount of butadiene equal to 74-84% of theoretical. The decomplexed CuCl was found to have excellent large pore porosity and good activity and capacity.

EXAMPLE 18.SLURRY OF C.P. BUTADIENE PLUS WATER WITH B&A SEGREGATED CU- PROUS CHLORIDE One liter of water was cooled to 5 C. 400 g. of Baker and Adamson segregated cuprous chloride were then added and butadiene was bubbled into the flask under the surface of the water with continuous stirring for a period of two hours. The stirrer was then stopped and the flask was allowed to stand overnight after which the cuprous chloride complex was filtered off, washed with alcohol and dried with ether in the usual manner. Upon dissociation the particles were found to have the desired large pore, porous structure and high activity and capacity in fluid bed tests.

EXAMPLE 19.fiLURRY OF C.P. LIQUID BUTADI- ENE WITH CUPROUS CHLORIDE RECRYSTAL- LIZED FROM HCl A sample of the cuprous chloride recrystallized from concentrated HCl (added to water) was covered with liquid C.P. butadiene (condensed in the reaction flask at 60 C.). The butadiene-cuprous chloride slurry was stirred for about one hour and finally the butadiene was boiled olf by warming the complex to room temperature. A dry powdery product was obtained which had com- 23 paratively low large pore porosity and comparatively poor activity and capacity.

EXAMPLE 20.COMPARISON OF THE PHYSICAL PROPERTIES AND FLUID BED PERFORMANCE OF THE CuCl AND CuBr PREPARED IN PRE- CEDING EXAMPLES The present example presents a comparison of the physical properties and performance in fluid bed tests on crude butadiene from steam cracking of the CuCl or CuBr prepared in the preceding examples. The fluid bed tests were conducted in a 27.5 inch length vertical glass reactor tube 28 mm. in diameter contained in a glass heating cooling jacket. Approximately 80 to 140 (in various runs) gms. of the CuCl or CuBr to be tested was charged (usually as the complex to prevent any deactivation by water in the air) along with approximately 100 gms. of Scotchlite glass beads to the reactor tube at room temperature. The reactor was then heated to approximately 90 C. while fluidizing with nitrogen to dissociate the complex (fluidized bed depth was 1220 inches). After no more butadiene was detected in the tail gas (measurement by gas chromatography) the supply of nitrogen was cut off and measurement was made of the amount of material present in the original complex (by condensing the total effluent gases from the fluid bed in a Dry Ice trap during the dissociation step).

The reactor tube was then cooled to C. and crude butadiene from steam cracking (31 wt. percent butadiene, 52 wt. percent isobutylene plus butene-l, the remainder being mainly paraflinic but also containing other butenes and acetylenes) was charged for one hour (temperature of the bed was about 10 C. above the temperature of the cooling jacket during the first 1530 minutes for the good materials tested). The activity of the CuCl or CuBr tested (butadiene present in the feed recovered) was determined by difference between the butadiene content of the feed and tail gases measured during the first 15 minutes of the run (additional analyses showed that this activity was essentially constant to capacity which was reached for the good materials in the first 15 to 30 minutes). Following absorption, i.e. charging of crude butadiene for one hour the bed was purged with 0.3 ft. of nitrogen at 0 C. to remove unreacted materials and then the bed was heated to 80 C. to effect decomplexing (no extraneous fluidization gases used except near end of the period when nitrogen was again supplied to make sure desorption was complete). The butadiene decomplexed was recovered in a Dry Ice trap and measured to determine the percent of theoretical capacity reached (theoretical capacity is 1 mol of butadiene for 2 mols of CuCl or CuBr). It is noted that the above described absorption and desorption steps were repeated at least 3 times and that the activity (percent recovery) and capacity (percent of theoretical) results reported in the table presented below are for the third cycle (similar resultswere obtained for the 1st cycle where such measurements were made). The absorption gas residence times reported were calculated from the gas flow rates assuming that the gas volume in the fluid bed was equal to the difference between the volume of the fluidized bed and the volume of the slumped bed.

EXAMPLE 21.-COMPARISON OF THE PHYSICAL PROPERTIES OF FLUID BED PERFORMANCE OF THE CuCl/1ND CuBr PREPARED IN PRECEDING EXAMPLES Fluid Bed Opera- Pore Volume, cc./g.

Orig. Gas tion 3rd Cycle Compl., Res. on Butad. Particle C014 Hg Porosimeter Fluid- Percent Time, Size, s Pr. ization of Theo. Sec. Cap, Ree, 800 A. 70-550 A. 550- 10,000 A.

Percent Percent 0,000 A,

I-lCl soln. of Bakers (3.1. (06.5%) CuCl- 100 1. 4 53 93 5/50 0. 02 G 5,

Buti'idicne complex added to water. Ex. Uneomplexed CuCl Reerystallizcd from 0 1. 7 3 14 10/50 4 Nil 4 0. 006 4 0. 009 4 0. 187 Do.

HGl (98.5100.0%). Ex. 1B. Evap. Butene-l soln. of 136m 0.1. 100 1.2 10 8 2/15 Nil 0.005 0.003 0.141 Very poor.

(98E0%) 8x101 complexed with buteue- 1. x. 1 Water added to H01 soln. of Bakers 100 1.7 79 78 100-300 .01 0.007 0.084 0.611 Excellent.

0.15. (90.5%) CuCl-butadiene oomplcx. Ex. 2. Acetylene added to L101 50111. of reagent- 35/200 0. 05 0. 008 0.018 0. 038

grade C11 0 precipitates complex directly. Ex. 4. HBrsolmot B&A 0.1. CuBr-buta-dieno 1.3 56 19 30/80 0.04 0.009 0.117 0.083 Good.

complex added to water. Ex. 5. N21250:) water soln. added to butediene 99 9 78 93 10/100 0. 01 0.002 0. 098 0. 198 Do.

purged water soln. of CuClz- Ex. 7. S0 and butadicne gases added to water 9 59 80 25/180 Nil 0.007 0. 197 0. 044 D0.

soln. oi CuGl Ex. 8. 1120 added to H01 soln. of Bakers C.P. 0. 9 0O 83 20/100 0. 01 0. 004 0. 059 0.061 Do.

(516.5%) C-uCl-acetonitrile complex.

x. 9. H O added to H01 soln. of Bakers 0.1. 100 0. 8 52 22 200/600 0.02 0.005 0. 009 0.135 Very poor.

(30.5%) CuCl-acrylonitrile complex.

x.1 H2O added to 1161 soln. of BdzA 0.1. 75 1.6 47 82 100/300 5 0. 01 5 0. 005 5 0.125 5 0.045 Excellent.

(98.6%) CuCl-carbon monoxide complex. Ex. 11. 11 0 added to 101 soln. of 36:11 (LP. 10/80 0.01 0.004 0.008 0.051 Good.

(928.0%) CuCl-ammonia complex. Ex. 11 11 0 added to H0] soln. of 1381A C.P. 90 0.8 62 82 100/400 0. 01 0.005 0.159 0.022 Excellent.

(98.6%) CuCl-methyl acetylene complex'. Ex. 12. Slurry of 0.1. Butadiene (10 vol. per- 76 1.7 67 94 5/00 Nil 0.000 0. 093 0.537 Good.

cent methanol) with BdzA segregated (100.4%) CuCl. E1. 13. Slurry of 0.1. Liquid Propylene with 2 1-2 8 21 15/90 Nil 0.006 0.000 0.028 Do;

CuCl Recrystallized from H01 (98.5- 4 15/90 4 Nil 4 0. 004 4 0.003 4 0. 050 100.0%). Ex. 13A. Slurry 0t Bakers C.P. (90.5%) 01101 18-32 20/100 Nil 0.003 0.020 0.100

with cone. HCl continuously purged with (3.1. butadiene. Ex. 13B. Slurry of C.P. Butadiene (40 vol. per- 92 5/50 0.001 0.005 0.190 0.257

cent methanol) with BdzA segregated (100.4%) CuCl. Ex. 14. Slurry of 0.1". Butadicne (73 vol. per- 37 20/90 0.01 0.002 0.060 0.201

cent methanol) with Bakers C.P. (96.5%) CuCl. Ex. 15. Slurry of 0.1. Butadiene (Plus metha- 07 1.8 64 32 5/50 0.01 Very poor.

1101) with H01 Recrystallized CuCl mostly See footnotes at end of table.

EXAMPLE 21Continued Fluid Bed Opera- Pore Volume, ce./g. Orig. Gas tion 3rd Cycle CompL, Res. on Butad. Particle C01 Hg Poroslmeter Fluid- Percent Time, Size, p Pr. ization of Theo. Sec. Cap., Ree, 800 A. 70-550 A. 550 10,000 A.

Percent Percent ,000 Slurry of Crude Liquid Butadiene with 74-84 2. 3 56 97 10/75 0. 01 0. 003 0.138 0. 294 Good.

CuCl recrystallized from H01 (98.5- 100.0%). Ex. 17. Slurry of C.P. Butadiene plus water 72 2. 8 45 64 10/30 0. 01 0. 003 0. 109 0. 271 Poor.

githlBdzA segregated (100.4%) CuCl.

X. Slurry of 0.1. liquid butadiene with 76 3.2 26 97 Nil 0.003 0.020 0.100 Fair.

CuCl recrystallized from H01 (98.5 100.0%). Ex. 19.

1 Theoretical complex is 2 mols C1111 mol complexing material except olefins and 1:1.

2 The activity of the 01101 remains essentially constant throughout the solid residence time until the capacity is reached and then drops to zero.

This example shows:

1) that repeated cycles use of 98.5100.0% purity CuCl on crude butadiene (Ex. 1B) does not produce the high activity (recovery percent) and high capacity obtained with the present new CuCl nor the high large pore porosity (55 0/ 10,000 A.) responsible for such activity and capacity.

(2) that use of monoolefins or ammonia as the complexing material (Ex. 1C and Ex. 13A) also does not produce the said high activity and capacity and high large pore porosity. Also, ammonia as the complexing material does not produce the high large pore porosity needed for high activity and capacity.

(3) that in the slurry preparations even with essentially pure (98.5100.0%) CuCl, high capacity activity 3 and accompanyin high large pore porosity are not obtained with the liquid diolefin e.g. butadiene alone (Ex. 19), i.e. in the absence of a monoolefin solvent, alcohol, glycol, or water activating material.

(4) that in the slurry preparations 99+% purity CuCl (as compared to low purities) produces higher activity, capacity, and large pore volume even with alcohol (methanol) which provides the best activation of the activating materials (Ex. 14 vs. Ex. 15

(5) that in the slurry preparations water is not as good an activating material as either monoolefins (mixed butenes in crude butadiene) or as alcohols (methanol) but that it does produce a very good material on all counts as compared to the poor materials described above. (Ex. 18, Ex. 17, Ex. 14.)

(6) that average particle size above about l0,u is necessary for good fiuidization even with fluidization aid glass beads added and that poor fiuiclization besides causing high losses of particles, etc. also badly affects recovery percent due to bypassing of vapors even where the 10,u particles have good 550/ 10.000 A. porosity.

It is noted that with respect to the CuCl purities given above that special analytical procedures involving blanketing with CO at all stages to prevent oxidation of CuCl to CuCl were required to obtain the precision analyses necessary to define the present invention. For example B&A C.P. cuprous chloride was found to be 98.6% CuCl vs. greater than 95% CuCl reported by the company presumably using A.C.S. standard procedure for analyzing reagent grade CuCl (CO blanketing not prescribed).

It is noted that in all the above preparations large pore 550/10,000 A. pore volume correlated closely with the activity (Recovery percent) and capacity measured in fluid bed tests. The capacity figure should first be noted and then the activity figure with due allowance being given for poor fluidization due to small particle size reducing the activity shown. It is further noted that although higher large pore (550/10,000 A.) porosities than about 0.039 cc./gm. i.e. about 10% are preferred, that a great improvement is obtained even with 10% large pore po- 3 Even after repeated complexing-decomplexing cycles, the capacity never exceeded 18%.

4 Measured after multiple use in fluid bed.

5 Pore volumes from similar sample prepared using Baker's C.P. CuCl rosity. It is also noted that much of the above data is on smaller particles than the optimum large spherical macroparticles and that it would be expected that the high large porosities preparations would be particularly advantageous since some of the porosity would be lost in the joining of microparticles. Additional experiments were conducted to determine the effect with low purity CuCl (Mallinckrodt analytical reagent gradeanalysis by present inventor 87.5% CuCl) in the precipitation of the complex from concentration HCl. It was found that the growth of the complex crystals effected essentially complete purification. The cuprous chloride removed from the complex upon dissociation was found to be greater than 99.5% and had a 550/10,000 A. pore volume of .106 cc./gms. It was further found that additions of copper pellets to the acid solution before the complexing reaction to reduce cupric copper as done in most of the preparations was not essential i.e. purity and 550/ 10,000 A. pore/ volume after dissociation of this complex were 99.5% CuCl and .096 cc./gms.

EXAMPLE 22.EFFECT OR RATE OF PRECIPITA- TION UPON SIZE OF ACTIVE PARTICLE Presented below is a comparison of the eifect upon particle size of the rate of precipitation or growth of the particle from solution. It is noted that all the preferred techniques for obtaining the said slow rate of crystal growth which produce the desired relatively large particles (fully discussed in this specification prior to the examples and used in the example preparations) all are encompassed within this single variable, i.e. precipitation rate. The precipitation rate is expressed in gms/hour/liter of solution containing the dissolved copper salt. This rate therefore takes into account the following variables which increase particle size: low concentration of dissolved copper salt, slow rate of addition of precipitating material, slow rate of increase in supersaturation (i.e. either addition of antisolvent, e.g. water, to the copper salt solution rather than vice versa, or alternatively addition of the copper salt solution to a limited amount of the antisolvent so that although initial precipitation as very fine crystals occurs (slight cloudiness) these are partially redissolved as the amount of solvent in the total mixture of solvent and antisolvent slowly increases and slow growth occurs (large particle precipitate) due to the simultaneously slow increase in the amount of CuCl or CuBr complex in the total mixture. The rates presented below were calculated in general as follows: the grams of complex recovered was divided by the time (hours) over which precipitation occurred and this figure was divided by the total liters of original cuprous salt solution. (Except when adding the acid to water, the time was assumed to be 3 seconds because the material precipitated as fast (e.g. 3 seconds) as the acid solution was added, due to the essentially complete insolubility of the complex.)

Precipitation Particle Preparation Rates, Size, n

gum/inn] liter Butadiene Complexes:

H01 Soln. of CuCl-Butadiene Complex Added to water. Ex. 1 -300 5/50 Water Added to H01 Solution of CuCl-Butadiene Complex. Ex. 2 77 100/300 NazSOa Water Solution Added to Butadiene Purged Water Solution of CuCl. Ex. 7 l00 10/100 S02 and Butadiene Gases Added to Water 10 Solution of CllCiz. Ex. 8 50 25/180 N o Complex-HC1 Solution of CuCl (uncomplexed) Mixed with water in Pipe 8, 000 1-5 Other Ligands-(Water added to HCl solution of Bakers 0.1. (96.5%) CuCl-Ligaud Complex):

Acetylene Ligand 680 50/150 Methylacetylene Ligand 700 40/150 Piperylene Ligand (added as gas diluted with N2 760 /100 Allene Ligand... 1,000 50/150 Carbon Monoxide Ligand 680 50/300 EXAMPLE 23.FLUID BED RECOVERY OF ETHYLENE 2O USING VARIOUS CuCl PREPARATIONS Fluid Bed Results with 50:50 Ethylene/Ethane sgln. oi 9 wt. percent CHaOH in Operating temp. length was normally 1 to 1.5

of fluid bed was 35 C. at 1 atmosphere. Bun hrs. on feed which was much longer than needed to saturate the bed. Then the bed was purged with about 10 liters 35 C. and finally heated to 80 C. to recover the O2.

b Initial ethylene recovery is reported since the recovery decreases steadily throughout the active life of the bed in a given run.

EXAMPLE 26.ADDITIONAL COMPLEXING AGENTS FALLING WITHIN THE CLASS OF THOSE PRODUCING ACTIVE CuCl OR CuBr The following solid complexes in addition to those described above and in the literature which produce stable complexes having a ratio of copper to complexing agent above 1:1 were prepared in the laboratory (all 2:1 or above): With CuCl: ethyl acetylene, vinyl acetylene, 1,5 hexadiene, HCN, cyclopentadiene, cyclooctadienes (1,3; 1,4; and 1,5 all prepared) cyclododecatriene; with CuBr: acetylene, methyl acetylene, allene, piperylcne. These complexes were prepared by absorption of the complexing material on solid CuCl or CuBr and thus did not produce the active porous material. However, this example indicates that these complexing agents can be used to prepare the active CuCl or CuBr by use of the techniques of the present invention.

EXAMPLE 27.-BUTADIENE RECOVERY IN A PRESSURIZED FLUID BED Crude butadiene from steam cracking was supplied to a 2" LD. 15 ft. in length glass reactor containing a 2- pass bundle of /2" diameter cooling-heating tubes extending 10 ft. vertically through the fluid bed. The temperature in the fluid bed was measured by thermocouples and the amount of CuCl (prepared by precipitation of the butadiene-CuCl complex in Water, average particle size 1060,u) was varied to determine percent butadiene recovery with different contacting times. The following results were obtained.

EXAMPLE 24.-ADDITIONAL SYSTEMS SEPARATED 0N FLUID CuCl BEDS Complexlng Complexing Product Recovered System (percents by wt.)

50 Ethylene-50 Ethane 50 Allene-EO Methylacetylenen l0 Acetonltrile-QO Acrylonitrile- 90 Piperylenes-IO Cyclopentene 92 Allene-8 Chloropropene 33 Butadiene-67 Mixed Butenes 70 Trnnsplperyleneels Piperyle 66 cis Piperylene-34 Cyclopentene 5 Butadiene-95 Isobutylene l CuCl prepared by precipitation of H01 soln. of CuCl-Butadiene Complex. (Ex. 1.)

1 99+% pure.

EXAMPLE 25.-.ADDITIONAL SYSTEMS SEPARATED 1N FIXED CuCl BEDS Complexing Oomplexing Product Recovered System (percents by wt.) Temp, Pressure, in High Purity 2 C. atmos.

50 Ethylene-50 Ethane (l25 16-20 Ethylene.

43 Ethylene-48 Ethane-9 Propylene l8 20 Do.

65 cls Plperylene- Cyclopentene 22 1 are Piperylene.

06.5 trans Piperylene-3.5 Cyclopentene 22 1 trans Piperylene.

00-55 Ethylene 0 20 C 45 Butadiene- Mixed Butenes 0 2. 5 Butadiene.

93 Allelic-7 Chloropropene. 0 1 99.9 Allene.

98.3 Methylacetylene-Impurities 0 1 90.0 Methylacetylene l CuCl prepared by precipitation of HCl soln. oi CuCl-Butadiene Complexv (Ex. 1.)

Total Aver. Height of Super. Tail Gas, Percent Cycle Pressure, Temp. of Fluid Salt Velocity, M01 Per- Butadiene Number p.s.i.g. Cogrfiplex, Bed, Ft. f.p.s. cent 0411 Recovery 4. 01 40 ca. 115 2. 5 0.30 22. 3 41. 7 4. 02 40 ca. 115 2. 5 0.30 24. 2 35.2 4. 04 40 113 6. 0.30 14. 9 64. 4.05 40 112 5. 7 0. 30 15. 3 63. 3 4.06 40 114 5.0 0.30 18. 3 54. 5 4.07 40 117 4. 5 0.30 19. 3 51. 4 4. 08 40 110 9.0 0.30 4. 7 90. 0 5.05 40 105 9. 1 0.30 6. 2 86. 6 5. 06 40 108 9. 5 0.30 6. 3 86. 4 5. 07 40 112 9. 6 0. 30 9. 3 79. 2 5. 40 112 10. 0 0.30 6. 6 85. 7

EXAMPLE 28.ADDITIONAL COMPLEXING AGENTS PRODUCING ACTIVE POROUS CuCl AND THEIR EFFECT ON CRYSTALLIZATION TECHNIQUE saturated solution of Bakers C.P. CuCl in concentrated HCl at 23 tested and then 375 ml. of N2 purged distilled water was added while of ligand. The crystals were then filtered ofi, washed with isopropanol and ether, and finally dried Preparation: 125 ml. of a water white saturated with the ligand being C. was filtered and then continuing the addition under a flowing stream of the ligand.

Ligand in Pore Volume After Dissociation of Compd. Complex, Ligand in Particle Precipi- Ligand Weight Complex, 0 C14 Mercury Size, tation Percent Percent 0-80 Microns Rate, of CuCl oi Theor. A. 70-550 55010,000 10,000 g./l./hr.

A. A. A.

Acetylene- 6. 8 52 0. 01 0.043 0.024 0.002 50/150 680 Methylacetylene 17. 9 90 0. 01 0. 005 0. 159 0. 022 50/150 760 Gaseous Piperylene in N g 32. 9 97 0. 01 0. 006 0. 098 0. 367 /100 760 Allene 20. 0 100 0. 01 0. 009 0. 168 0. 009 50/150 1, 000 Carbon Monoxide- 20. 0 72 0. 01 O. 005 0. 125 0. 045 50/ 300 680 Ammonia 0. 7 2 0. 01 0. 004 0. 008 0. 051 20/100 600 e 85% trans-15% cis piperylene mixture. b Based on 2N H3 per CuCl.

This example indicates that acetylenes, conjugated and nonconjugated diolefins, and carbon monoxide are excellent complexing agents to produce the desired high activity, high large pore porosity cuprous chloride and cuprous bromide complexing agents. It also indicates that ammonia is not a suitable complexing agent and further indicates that 0 acetylenes are much better complexing agents than acetylene. Thus acetylene is known to behave in a considerably different way in chemical reactions, e.g., higher activity as compared to the higher acetylenes.

This example also indicates that higher allowable precipitation rates to obtain large particle size and efiicient fiuidization may be used with these complexing agents as compared to butadiene because the cuprous chloride or cuprous bromide complex with these materials is more soluble in the dilute acid than are the butadiene complexes.

EXAMPLE 29.-COMPLEXING NEAR THE DEW POINT WITH ACTIVE CuCl Com- Butadiene F. Cycle Diluent plexing Partial From Capacity, No. Time, Pressure, Dew Percent Min. Atm Point 13 Nitrogen 12 0.30 57 2O 6 Butane 12 0.37 5 35 8 Butane-1 8 0.37 10 45 23 Isobutylene 12 0.31 12 52 3. Refinery Stream. 13 0.35 1 12 47 Approximate.

The above data indicate the importance of operating within F preferably within 12 F., of the dew point both with and without monoolefin solvent activators; i.e., the latter additionally improve capacity as can be seen from the butane run.

Similar data were obtained on CP. butadiene, complexing time 1 hour (heat being given oif, indicated complexing substantially complete by this time) at various temperatures.

Cycle Number Complexing F. from Capacity Temperature Dew Point This data indicates the same results are obtained by varying temperature as by varying partial pressures. It also indicates that capacity can be restored by returning the conditions near the dew point. It is noted that activity appeared to be proportional to capacity in all the cycles reported in this example.

EXAMPLE 30.LIQUID PHASE ETHYLENE RECOVERY 560 grams of commercial (Bower C.P.) cuprous chloride and 560 grams of the same commercial cuprous chloride activated by liquid phase complexing with a 35 Wt. percent butadiene, 65 Wt. percent 0.; monoolefin mixture and vapor phase dissociating were separately reacted in a stirred autoclave with a liquid mixture of approximately 2.5 mol percent ethylene dissolved in n-heptane, i.e., Raoults law calculation from partial pressure ethylene in the exit gas (eflFect of heptane partial pressure in vapor phase is negligible). The data are:

Activated Commercial CuCl uCl Feed Gas:

Ethylene, mol percent 41. 6 47. 9 Propylene, mol percent 9. 1 Nitrogen, mol percent. 49. 3 52. 1 Feed Gas Rate, ftfi/hr 12. 25 12.25 Slurry Concentration, wt. percent 50 50 Temperature, F -30 30 Pressure, p.s.i 200 200 Equilibrium, partial pressure ethylene in exit gas, p.s.i.a 5.8 6.2 Complexing Time, Hr l 1 Percent of Theoretical solids loading:

Ethylene 52. 5 30. 0 Ethylene plus propylene 57. 5

In this specification and in the appended claims the terms absorption and desorption are used in their

Images